Stroke volume variation and pulse pressure variation measured at pulmonary arterial level versus pulse pressure variation measured at systemic arterial level: an exploratory study


Fluid therapy; stroke volume; pulmonary artery; hemodynamic monitoring; cardiac output

Published online: Apr 21 2022

M. Slabbaert, M. Vandenheuvel, P. Wouters

Department of Anesthesiology, Ghent University Hospital, Ghent, Belgium.


Problem: Optimized fluid management reduces morbidity and mortality in major operations, including cardiac surgery (1, 2). Validated dynamic parameters were developed to predict fluid responsiveness but have limitations (3-5). This study explores the potential added value of stroke volume variation and pulse pressure variation measured at pulmonary arterial level in predicting fluid responsiveness (6).

Methods: In twenty adult patients undergoing coronary artery bridging and/or aortic valve replacement, the systemic and pulmonary arterial (SA and PA) pressure curves were stored using Datagrabber software. A custom-made MATLAB code automatically analyzed the pulse pressures, from which the SA and PA pulse pressure variations (PPV_SA and PPV_PA) were calculated. Simultaneously, stroke volumes in the pulmonary artery were measured and PA stroke volume variation (SVV_PA) was calculated. These three parameters were determined in both closed and open chest situations and further analyzed using RStudio software.

Results: No correlation could be established between SVV_PA and fluid responsiveness as defined by PPV_SA above 12% (3). In the case of PPV_PA, a mild correlation was seen with the above-mentioned fluid responsiveness, with an R² of 25%. Bland Altman analysis of PPV_PA and PPV_SA showed proportional bias with broad limits of agreement. ROC curve analysis attempted to determine a new cutoff value for PPV_PA, which was 33% (sensitivity 100%, specificity 86%).

Conclusions: SVV_PA could not be correlated with fluid responsiveness as previously defined. PPV_PA appears to be mildly correlated with fluid responsiveness and has a relatively faster increase in variation than PPV_SA. Therefore, a new fluid responsiveness cutoff for PPV_PA is proposed. Further research is needed to confirm these findings and to explore further the sensitivity and specificity regarding fluid responsiveness.